A model of noise that results from changes in cell-cycle–stage distributions in the population at different growth rates predicts global changes in noise across conditions. (A) Our model prediction of how noise (CV2, y-axis) globally changes with the G2 fraction of the population (x-axis), according to Equation 1, with C = 0.065 (black line). Colored points represent experimental data. The fraction of G2 cells in the population at each growth condition was determined by flow cytometry measurements of a histone-GFP reporter construct (GFP-HHF1) (B, x-axis). The extrinsic noise limit for the batch-growth conditions (glucose, glucose w/o AA, galactose w/o AA, and ethanol) was extracted by a linear fit to the ungated promoter–reporter measurements (Supplemental Fig. S9, y-axis). The extrinsic noise limit for the chemostat condition was computed as the average CV2 of the three promoter–reporter strains assayed in the chemostat (RPL28, HHT2, and TDH2). (B) The fraction of G2 cells in the population increases with growth rate, as determined by flow cytometry measurements of GFP-HHF1. (C) In addition to the measurements of the entire library in different batch conditions, three high-expressing strains from the library were grown in a His-limited chemostat (dilution rate of 0.04h−1). Shown are their CV2 values (y-axis) versus the fraction of the population in G2 (x-axis) in all assayed conditions. Lines are cubic interpolations between measured points and serve as a visual guide.
